UPDATING THE ESTIMATE OF THE SOURCES OF PHOSPHORUS IN UK WATERS - White and Hammond (Defra Funded)

Executive Summary

  • This desk study aimed to identify current sources of phosphorus (P) loads to UK waters and compare the updated P source apportionment with previous apportionments in the UK and other European countries.
  • The study reviewed methods for estimating and apportioning P loads, suggesting a combination of industrial point source inventories and diffuse P load estimates.
  • Diffuse total phosphorus (TP) loads were calculated using standard export coefficients, land cover data (Land Cover Map 2000), agricultural stocking densities (2004), and human populations (2001).
  • Maximal point source industrial P loads were obtained from the Environment Agency’s Pollution Inventory. Soluble reactive phosphorus (SRP) loads were estimated from TP loads using specific SRP:TP ratios.

Findings:

  • The TP load to waters of England, Wales, and Scotland was estimated at 41.6 ewline kt y^{-1}.
    • Agriculture contributed 11.8
      ewline kt y^{-1} (28.3%).
    • Households contributed 25.3
      ewline kt y^{-1} (60.7%).
    • Industry contributed 1.9
      ewline kt y^{-1} (4.6%).
    • Background sources contributed 2.7
      ewline kt y^{-1} (6.5%).
  • Using a higher value for human point source contributions (STWs) of 0.61 kg TP per capita, the proportions change:
    • Agriculture: 22.5%
    • Households: 68.7% (34.8
      ewline kt y^{-1})
    • Industry: 3.6%
    • Background sources: 5.2%
  • The SRP load to waters was estimated at 31.3 ewline kt y^{-1}.
    • Agriculture contributed 5.8
      ewline kt y^{-1} (18.6%).
    • Households contributed 21.1
      ewline kt y^{-1} (67.4%).
    • Industry contributed 1.7
      ewline kt y^{-1} (5.5%).
    • Background sources contributed 2.7
      ewline kt y^{-1} (8.5%).
  • Area normalised TP load:
    • Agriculture: 0.55
      ewline kg TP ha^{-1} y^{-1}
    • All sources: 1.94
      ewline kg TP ha^{-1} y^{-1}
  • Phosphorus loads varied widely between River Basin Districts (RBDs).
    • Heaviest TP loads: Humber, Thames, and Severn.
    • Lightest TP loads: Northumbria, Solway Tweed, and Dee.
  • Area-normalised annual TP load:
    • Highest: Thames (4.41
      ewline kg ha^{-1} y^{-1}), North West (3.37
      ewline kg ha^{-1} y^{-1}), and Humber (3.33
      ewline kg ha^{-1} y^{-1})
    • Lowest: Anglian (1.39
      ewline kg ha^{-1} y^{-1}), Scotland (0.84
      ewline kg ha^{-1} y^{-1}), and Solway Tweed (0.57
      ewline kg ha^{-1} y^{-1})
  • Agricultural contributions to TP loads of RBDs:
    • Highest: Western Wales (62.4%), Solway Tweed (51.9%), and Severn (48.2%)
    • Lowest: Thames (9.7%)
  • Industrial Point Sources:
    • Significantly high contribution to P load in Scottish waters (23.4%).
  • Household Sources:
    • Large contributions (>70%) to P loads in South East, North West, and Thames RBDs.
  • England:
    • Contributes the most TP to UK waters.
    • Agriculture contributes:
    • 26% of the English TP load
    • 22% of the Scottish TP load
    • 57% of the Welsh TP load
    • Approximately 47% of the TP load to inland and coastal waters of Northern Ireland.

Objective 01. Methods for Estimating P Loads and Source Apportionment

  • Literature survey to identify methods for estimating P loads and source apportionment to inland and coastal waters.
  • Assessment of methods to determine catchment-scale P loads and apportion these loads to different P sources.
  • The simplest, most versatile, and intuitive method is to sum the contribution from:
    • Point sources (from national inventories).
    • Empirical models using region-specific export coefficients for different land cover classes.
    • Data for agricultural stocking densities and human populations.

01.01 Methods for Estimating P Loads

  • Methods separate into:
    • Data from river flow and water quality monitoring programs.
    • Calculating P loads from inventories of point P sources plus estimates of diffuse P sources.

01.01.01 Estimating P Load Using Data from River Flow and Water Quality Monitoring Programs

  • Common methods advocated by the International Oslo and Paris (OSPAR) Commission.
  • River loads are calculated as the sum of the products of instantaneous flow and concentration measurements.
  • Databases:
    • National River Flow Archive (NRFA): daily mean flows.
    • Harmonised Monitoring Scheme (HMS): concentrations of orthophosphate and TP.
  • UK data for orthophosphate is “broadly reliable,” but TP data has “significant, substantial or severe problems”.
  • 'Instantaneous monitoring’ approach used for estuaries and smaller catchments.
  • Four methods to estimate annual loads from river flows and measured concentrations:
    • Product of annual mean flow and annual mean concentration.
    • Sum of the products of flow and concentration data sampled regularly throughout the year.
    • ‘Flow weighted’ version of the second method (preferred).
    • Parameterized empirical relationship between load and flow rate.
  • Estimating TP loads requires high-frequency sampling during storm events.
  • Calculating TP loads based on monthly or fortnightly sampling can misestimate the real load considerably
  • Accuracy and precision of estimated P load influenced by:
    • Equations used to estimate loads.
    • Regularity and frequency of sampling.
    • Method used to estimate water flow.
    • Chemical methods used to estimate orthophosphate and TP concentrations.
  • It is often difficult to use monitoring data to estimate the absolute P loads to a large catchment unless the P retention properties of the catchment are known.
  • Retention is a collective term for the diverse biological, geochemical and hydrological processes that remove, transform, or retard the transport of P within a watercourse.
    • It includes processes such as sedimentation or settling of particulate P and the uptake and storage of nutrients by plants, algae, microbes and fish.
    • It is a seasonal phenomenon, with P being retained during the summertime and periods of low water flow and P being remobilised during periods of high water flows in the autumn and winter.
  • Riverine retention of P in the summer may be considerable (over 50% of the diffuse P load to a watercourse).
  • An estimate of the annual P retention within a watercourse can be calculated from the empirical relationships derived by Behrendt and Opitz (2000), who suggest that P retention is a function solely of specific runoff to a river basin.

01.01.02 Estimating P Loads From Inventories of Point Sources and Estimates of Diffuse Sources

  • Point source household and industrial P loads can be obtained directly from monitoring data of industrial sources and wastewater treatment plants (STWs).

  • Upper value for industrial discharges directly to controlled waters can be obtained from the Environmental Agency’s Pollution Inventory.

  • Greatest industrial discharges arise from chemical factories, abattoirs, creameries and food processing.

  • When data for the actual point source P load from STWs is unavailable, household P loads are estimated indirectly from population census data, assuming an average per capita P load after treatment by STWs.

  • Calculation of the mean TP load divided by the ‘population equivalent’ (PE) served by the largest 62 STWs of Severn Trent Water listed on the EAPI suggests a value of 0.42
    ewline kg TP PE^{-1}y^{-1}

  • Assumed to be derived from human excreta (66%) household waste (28%) and detergents (6%)

  • STWs operating either biological and/or chemical treatments (tertiary treatment) can remove more than 80% of the incoming P.

  • Alternatively, the P load can be calculated from the product of an assumed constant P concentration STW effluent (10
    ewline mg l^{-1} in the UK) and the total permitted daily volume of effluent for the STW in each catchment.

  • Use of export coefficients may not distinguish between people connected to mains sewerage systems and those served by septic tanks.

  • The use of permitted volumes leads to an over-estimation of flux from STWs, but omits the contribution from septic tanks.

  • Septic tank systems are the main method of disposal of human waste in most areas of rural Britain.

  • The effluent from these tanks usually drains to a soakaway where the P is dissipated into the surrounding soil.

  • TP load from septic tanks approximates 0.24 to 0.40 kg TP person-1 y -1 and that from dwellings delivering effluent directly to rivers approximates 0.63-0.72 kg P person-1 y -1 .

  • Burns (2004) reported that protected horticulture contributed to the P loads of 604 river catchments in England and Wales and amounted to 594 t TP y -1, most of which came from vegetable and fruit crops (61%), cut flowers (11%) and nursery stock (21%).

  • For crops grown hydroponically, such as tomatoes, cucumbers and peppers, TP losses can amount to 330 kg P ha-1 y -1

  • Diffuse P loads arise from runoff from natural landscapes, agricultural fields and rural and urban surfaces, wastes from non-sewered populations and animals, atmospheric deposition and inputs from aquifers.

  • Diffuse loads can be estimated either:

    • By dividing the survey area into different land use categories using appropriate export coefficients.
    • From modelling the P input and transport processes occurring in a survey area.

  • Estimating P loads for diffuse sources from export coefficients

    • Export coefficients matched to specific descriptions of land usage
    • Simple method to estimate the average annual diffuse P loads within large, diverse catchments, using data from readily available databases.

Limitations to this approach:

  • Estimates of P loads depend critically on the choice of export coefficients for particular land-cover types and climatic conditions.

  • The method lacks spatial resolution and relies on average export coefficients for particular land uses.

  • The method is designed to produce estimates of annual P loads and is difficult in predicting P losses over shorter timescales.

  • Since export coefficients seek to integrate all the physical processes operating on P fluxes within the catchment, they do not afford much insight to the complex biological or physical dynamics of P fluxes in the environment.

  • Export coefficients must be tailored to a particular land use and topography.

  • The diffuse urban P load originates from a variety of sources including atmospheric emissions, runoff from roads, driveways, roofs, parking lots, construction sites and gardens, which exhibit seasonality according to rainfall and leaf fall, urban litter, car- washing and industrial spills

  • Soluble P can contribute significantly to the diffuse urban P load

  • Diffuse P load from an animal is estimated as the product of the P in wastes voided by the animal, the amount of these applied to land, and an export coefficient to describe the movement of animal wastes to the watercourse.

  • The atmospheric P load to the land is estimated as the product of P deposited to an area and an export coefficient to describe its movement to a watercourse.

  • Estimating P loads for diffuse sources from mechanistic models

    • Mechanistic models seek to describe P dynamics in the environment using physical parameters
    • Field-scale models
    • Catchment-scale models

01.02 Methodologies for Estimating Source Apportionment

  • Methods are based on either:
    • Water quality monitoring data or
    • Inventories of point P sources plus estimates of diffuse P sources based on either export coefficients or mechanistic models.

01.02.01 Estimating Source Apportionment From Water Quality Data

  • The simplest method for estimating the contribution of diffuse sources to a P load is from the difference in the P load estimated from water flow and P concentration data and the point source inventory.
  • To address the criticism of neglecting the occurrence of retention or loss in the river system a coefficient can be included in the calculation to account for systematic losses.
  • Alternatively, source apportionment between point and diffuse sources, and/or between contrasting land uses, may be based on statistical analysis of observed P loads and explanatory variables
  • A complementary, but indirect, apportionment approach based on water quality data uses specific physical, chemical and/or biological properties as tracers for particular P sources.

01.02.02 Estimating Source Apportionment From Inventories of Point Sources and Estimates of Diffuse P Loads

  • Source apportionment of P is often accomplished by combining inventories of point sources and estimates of diffuse loads estimated from export coefficients or mechanistic modelling.

Objective 02. Quantifying P Loads and Source Apportionment for UK Waters

  • There is great variability in export coefficients derived for similar land usage.

  • To estimate diffuse TP loads for RBDs, we have used export coefficients adapted for the 72 Level-3 subclasses of land cover in the Land Cover Map 2000 (LCM2000); condensed to 27 LCM2000 Level 2 land cover classes.

  • The TP load is calculated from agricultural stocking densities, human populations and atmospheric deposition in addition to land use statistics.

  • It is assumed that 30% of the rainfall on land reaches a watercourse and that the TP load in rainfall is 0.4
    ewline kg TP ha^{-1} y^{-1}

  • The TP loads from human sewerage were calculated from the human populations in RBD multiplied by an estimated TP load per person per year.

  • This estimate was set to the mean TP load per ‘population equivalent’ (PE) served for the largest 62 STWs of Severn Trent Water listed on the EAPI.

  • SRP loads were determined from the TP load from LCM2000 land use classes, livestock, human and rainfall using estimates of their SRP/TP quotients

  • The areas of land use classes, including the area of inland waters receiving atmospheric deposition, in each RBD were derived from the LCM2000 data set, supplied under license by Defra GIS Department

  • Livestock numbers were obtained from agricultural census data for 2004 and human populations were taken from the 2001 Census data obtained from the Office of National Statistics.

Apportionment Results (Divided into Sectors):

  • Agriculture: Glasshouses, improved grassland, moors and heaths, arable cereals, field horticulture and livestock (cattle, pigs, sheep, poultry).

  • Human and household: Point (STWs) and diffuse urban sources.

  • Industrial: Point and diffuse sources.

  • Background: Orchards, woodlands, forests, wetlands (bog, fen, marsh, swamp) and atmospheric deposition.

  • Total TP loads for land use classes were calculated by multiplying their area and export coefficient values.

  • Diffuse TP loads from livestock were calculated by multiplying their number and TP load per capita.

  • Human point source (STW) loads were calculated by multiplying population statistics and TP load per population equivalent.

  • Phosphorus loads from industrial point sources releasing P into controlled waters were obtained from the EAPI and atmospheric deposition was calculated as the sum of the atmospheric P deposition to surface waters and 30% of the atmospheric P deposition to the land.

  • The TP load to England, Wales and Scotland was estimated to be about 41.6
    ewline kt y^{-1}.

    • The amount of TP apportioned to agriculture is 11.8
      ewline kt y^{-1} (28.3%), to households is 25.3
      ewline kt y^{-1} (60.7%), to industry is 1.9
      ewline kt y^{-1} (4.6%) and to background sources is 2.7
      ewline kt y^{-1} (6.5%).

  • If the higher value of 0.61 kg TP per capita is used for human point source contributions (STWs), as calculated from the load of all STWs, then these proportions change, with agriculture accounting for 22.5%, households accounting for 68.7% (34.8
    ewline kt y^{-1}), industry accounting for 3.6% and background sources accounting for 5.2%.

  • The SRP load to England, Wales and Scotland was estimated to be about 31.3
    ewline kt y^{-1}.

    • The amount of SRP apportioned to agriculture is 5.8
      ewline kt y^{-1} (18.6%), to households is 21.1
      ewline kt y^{-1} (67.4%), to industry is 1.7
      ewline kt y^{-1} (5.5%) and to background sources is 2.7
      ewline kt y^{-1} (8.5%).

Agricultural contributions (To the TP load from agriculture):

  • Improved grassland contributed 18.0%
  • Other grassland, moors and heath contributed 1.2%
  • Arable agriculture contributed 9.6%
  • Field horticulture contributed 16.8%
  • Glasshouses contributed 6.1%
  • Livestock contributed 48.3%

Livestock contribution (To the TP load from livestock):

  • Cattle contributed 37.5%
  • Sheep 13.2%
  • Pigs 33.0%
  • Poultry 15.0%
  • Horses 1.3%

Household contributions (To the TP load from households):

  • Diffuse urban TP contributed 4.4%
  • STWs contributed 95.6%

Other

  • Chemical industries contributed 75% and combustion processes contributed 10% to the industrial TP load.
  • The ‘average’ area normalised TP load to England, Wales and Scotland from all diffuse sources is about 0.83
    ewline kg TP ha^{-1}y^{-1}, from agricultural sources is about 0.55
    ewline kg TP ha^{-1}y^{-1} and that from all sources is about 1.94
    ewline kg TP ha^{-1}y^{-1}.

The TP loads to watercourses differ widely between RBD. Perhaps unsurprisingly, the heaviest P loads occur in the heavily populated and/or extensive RBD of Humber, Thames and Severn

  • Area-normalised annual TP load:
    • Highest: Thames (4.41
      ewline kg ha^{-1}y^{-1}), North West (3.37
      ewline kg ha^{-1}y^{-1}) and Humber (3.33
      ewline kg ha^{-1}y^{-1}).
    • Lowest: Anglian (1.39
      ewline kg ha^{-1}y^{-1}), Scotland (0.84
      ewline kg ha^{-1}y^{-1}) and Solway Tweed (0.57
      ewline kg ha^{-1}y^{-1}).
  • The agricultural contributions to the TP loads of RBDs followed the order Western Wales (62.4%), Solway Tweed (51.9%), Severn (48.2%), Dee (45.9%), South West (43.3%), Humber (34.9%), South East (21.8%), Anglian (20.35%), Scotland (18.6%), Northumbria (16.1%), North West (11.8%) and Thames (9.7%).
  • There is a notably high contribution from industrial point sources to the P load of Scottish waters (23.4%), and large contributions (>70%) from household sources to the P loads of the South East, North West and Thames RBDs.

Objective 03: Comparison of P Loads and Source Apportionment for England, Wales, Scotland and Northern Ireland.

  • TP loads for England, Wales and Scotland were estimated based on land use statistics, livestock numbers and human populations

  • Cross-boarder RBDs partitioned appropriately into respective national areas.

  • Data for apportionment of phosphorus sources for England were derived from values for the Anglian, Humber, North West, Northumbria, South East, South West and Thames RBDs and the areas of the Solway Tweed, Dee and Severn covering England.

  • Data for the apportionment of phosphorus sources for Wales were derived from the values obtained in Objective 02 for the Western Wales RBD and the areas of the Dee and Severn covering Wales.

  • Data for the apportionment of phosphorus sources for Scotland were derived from the values obtained in Objective 02 for the Scotland RBD and the area of the Solway Tweed covering Scotland

  • England contributes most to the TP load of UK waters

    • Agriculture in England contributes 26.0% of this TP load.
    • Household and Industrial diffuse and point sources together contribute 69.2% to the TP load of English waters.

  • Agriculture contributes less (22%) of the TP load of Scottish waters and more (57%) of the total TP load to Welsh waters.

  • Livestock and improved grasslands contribute 66% of the agricultural P loads of England, Wales and Scotland

  • Recently, Smith et al. (2005) estimated that agriculture contributed about 47% of all P exports to inland and coastal waters in Northern Ireland

Objective 04. An Appraisal of the Impact of P Pollution on the Aquatic Ecosystem in the UK, with Specific Consideration of the Impacts of Various P-Sources in Different Seasons and Sensitivities of Water Bodies to Eutrophication

04.01 Definition and Threshold P Concentrations for Eutrophication

  • Eutrophication is defined as “the enrichment of waters by inorganic plant nutrients which results in the stimulation of an array of symptomatic changes. These include the increased population of algae and/or other aquatic plants affecting the quality of the water and disturbing the balance of organisms present within it”
  • The sensitivity of a water body to eutrophication depends upon climate (rainfall, light and temperature), water chemistry, residence time and depth
  • Shallow, standing waters are more susceptible to eutrophication than fast-flowing rivers.
  • Hard water lakes are generally more susceptible to eutrophication than soft water lakes, reflecting the greater fertility of base rich soils.
  • The trophic state of most lakes and rivers in the UK appears to be determined by P status.
  • Mainstone and Parr (2002) suggested pragmatic management targets of between 20 and 100 µg SRP L-1 for UK rivers, depending on river type, and the Environmental Agency Rivers Task Team suggested threshold values for ‘good ecological status’ of 40-50 µg SRP L-1 for non- calcareous rivers and 120 µg SRP L-1 for calcareous rivers

04.02 Sensitivity of UK Surface Waters to Eutrophication

  • Risk of eutrophication is now widespread in the UK and many inland waters contain excessive P concentrations derived from anthropogenic sources
  • Median TP concentrations < 30 µg TP L-1 for rivers with high altitude catchments
  • Median SRP concentrations exceeded100
    ewline µg SRP L^{-1}in 78 rivers values in excess of1000
    ewline µg SRP L^{-1}in 16 rivers.
  • Highest mean concentrations associated with rivers draining the most populated catchments and receiving major point-source inputs from sewage treatment works
  • 73% of the lakes in the Scottish Highlands was oligotrophic.
  • 69% of lakes in England and Wales had TP concentrations greater than100
    ewline µg L^{-1}only 8% had TP concentrations less than35
    ewline µg P L^{-1}
  • Recently, Heathwaite et al. (2005) applied an export coefficient model based on land use and animal stocking data to determine P concentrations in all lakes in Great Britain of more than 1 ha area (>14,000 lakes). Their analysis suggested that 51% of lakes in Great Britain (and 88% in England) were at risk of not meeting the “good ecological status” objective of the EU Water Framework Directive because of diffuse agricultural P loads.
  • Report indicated that 47.4% of total river length, 36.8% of lakes and 17.4% of groundwaters are at risk, or probably at risk, from diffuse agricultural P sources

04.03 Forms of Aquatic Phosphorus

  • Phosphate is present in watercourses in many forms that are arbitrarily divided into particulate P (PP) and dissolved P (DP).
  • Forests loose P in DOP and MRP forms, but rarely as PP.
  • Grasslands loose most P in soluble forms due to their dense vegetative cover, which prevents particulate losses.
  • Drained fields loose large quantities of both DP and PP.
  • Losses from arable land are occasionally dominated by DP forms also.
  • High P losses from land growing horticultural and other wide row crops since they require high P inputs, there is more soil disturbance and the soil is often left barren for longer periods
  • The particles transported in runoff normally have higher P concentrations than the soil from which they originated
  • In urban areas, 15 to 80% of the diffuse TP load is MRP and the P effluent from STWs is also predominantly MRP.
  • It has been estimated that 26% to 88% of the TP load of British rivers is associated with PP. This implies a large contribution of agriculture to the P loads of rivers. Furthermore, about 60 to 70% of the PP in rivers is defined as ‘non-apetite inorganic phosphate’, which contributes to the total bioavailable P

04.04 Seasonality of P Loading to Surface Waters

  • Eutrophication is linked primarily to P concentrations during periods of ecological sensitivity.
  • SRP concentrations in early spring are often found to be the best predictors of algal growth.
  • The P load from STWs is generally constant throughout the year, but the contribution of diffuse agricultural sources to instantaneous P loads is greater with increased hydrological activity
  • Thus, the spring and summer P loads from STWs make an immediate contribution towards eutrophication, whereas the winter agricultural P loads make a delayed contribution to eutrophication during the subsequent spring and summer.
  • The trend towards higher winter flows increases agricultural PP loads
  • From a study of water quality at 54 river sites in 7 UK catchments with diverse land use characteristics, Jarvie et al. (2005) observed that P loads from STWs rather than diffuse agricultural P loads posed the most significant risk for eutrophication, even in rural areas.

Objective 05. Comparing Updated P Loads and Apportionment Figures with Previously Published Data for the UK and Fellow EU Countries with Similar Climatic Conditions and Land Use.

05.01 Loads and Apportionment Studies for the UK

  • Several studies in the last 20 years have sought to determine the relative contribution of agriculture to the P loads to UK waters
  • Withers suggested that agriculture contributed about 36% to the P load of UK waters and Haygarth et al., using the AERC Export Coefficient Model, estimated that agriculture was responsible for 43% of the P load to UK waters in 1991
  • Defra suggested that agriculture was responsible for over 50% of the TP loads to UK surface waters in 2002
  • However, the source apportionment undertaken in here attributes only 28.3% of the total TP load to UK waters to agriculture (51.7% to land use and 48.3% to livestock), with about 60.7% attributed to human and household waste, 4.6% to industrial sources (point and diffuse) and 6.5% to background sources (atmospheric loads)
    A value of 22.5% is obtained for the contribution of diffuse agricultural sources to the total TP load of UK waters, If a higher export coefficient for human point sources is used in the calculation,
  • The TP load estimated in the present study is far lower than that estimated by Morse et al. (1993). This suggests a large reduction in TP loads since the 1990s.

05.02 Loads and Apportionment Studies for UK Coastal Waters

  • UK coastal waters divided into North Sea, English Channel, Celtic Sea, Irish Sea and Atlantic Ocean
  • The UK has supplied the OSPAR Commission annual mass loads of orthophosphate and total phosphorus to its coastal waters since 1990

05.03 Apportionment Studies of Specific UK Watersheds

  • Historical correlation between increased soil P and TP in watercourses provides circumstantial evidence that agriculture contributes significantly to riverine TP loads in the UK.
  • Estimates that 26 - 88% of the TP load to UK rivers is associated with PP , also suggest a significant contribution by agriculture to riverine loads
  • The diffuse (mostly agricultural) contributions has huge variation, reflecting contrasting land uses and human population density, ranging from 7% to 100%.
  • Dilution of TP with increasing water flows and the correlation of TP loads with boron, suggest significant contributions from smaller STWs and/or septic tanks even in rural areas
  • Diffuse P sources contributed most during periods of high flows, whilst point sources contributed significantly during periods of low flow.

05.04 Apportionment Studies for EU Countries

  • In 1993, it was estimated that about 50% of the P load to surface waters in the EU came from agriculture, with 41% from waste waters and 9% from background sources Values for the agricultural contribution to the P load of surface waters ranged from 43% for the UK to 73% for Ireland
  • Revised estimates were recently reported by Herbke et al. (2005) for the European Environment Agency, which suggest agricultural contributions to TP loads to surface waters of between 10% for Norway to 80% for Lithuania

Objective 06. A Discussion of the Updated P Source Apportionment for UK Waters in the Context of the Water Framework Directive (WFD) Article 5 Risk Assessment Maps and Data

  • Under the WFD, risk assessments for each RBD were conducted to establish the likely impact of human activity on surface and groundwaters within each RBD.

  • Surface waters within each RBD were classified into rivers, lakes, transitional waters (estuaries) and coastal waters.

  • These waters were also sub-divided based on natural factors, such as altitude, latitude, longitude, geology and size

  • No transitional or costal waters were deemed to be at risk from either diffuse or point P sources

  • Area-normalised data for the length of rivers ‘at risk’ from point P sources in each RBD did show a good correlation (r^2 = 0.9) with area-normalised annual TP loads

  • A positive correlation (r^2 = 0.32) was observed between river orthophosphate concentrations and area-normalised TP loads

  • These observations suggest a large contribution from point sources to bioavailable P and risk of eutrophication